South Korean professional Go player Lee Se-Dol puts his first stone against Google's Artificial Intelligence program, AlphaGo, in March 2016 in Seoul. He lost the five-match series. Google via Getty Images
South Korean professional Go player Lee Se-Dol puts his first stone against Google's Artificial Intelligence program, AlphaGo, in March 2016 in Seoul. He lost the five-match series.

Linux Foundation Spawns Child Foundation for AI

Deep Learning Foundation's official goal: promote open source innovation in AI, machine learning, deep learning

The Linux Foundation is throwing its hat into the artificial intelligence ring. On Tuesday, the open source organization announced the launch of the Deep Learning Foundation. The official goal is to promote open source innovation in AI, machine learning, and deep learning. Like most Linux Foundation projects, this mainly translates into the development of software -- in this case AI, machine learning, and deep learning applications.

Also like most of the foundation's projects, this one is hitting the ground running with a flagship software project, and sponsors that include Amdocs, AT&T, B.Yond, Baidu, Huawei, Nokia, Tech Mahindra, Tencent, Univa, and ZTE.

The software, which is already available for download, is the Acumos AI Project and was supplied by AT&T and India-based Tech Mahindra, a supplier of IT for telecoms. The platform makes it easy to create, share, and deploy machine learning, deep learning, and analytics models, either in stand-alone implementations or integrated into other apps. The Linux Foundation will host the project along with a supporting Acumos Marketplace, hoping to build a community of developers around the project.

"With the launch of Acumos, the vision to make AI accessible to a wider community of users across diverse industries and hence making it easy to implement AI applications are now a reality," Tech Mahindra's president of communications business, Manish Vyas, said in a statement. "The potent force of AI, data and cloud that the Acumos platform inherently acts upon will help realize the transformation journey that Tech Mahindra and its clients have embarked on."

Down the road, the Linux Foundation expects to see code contributions from Baidu and Tencent. Baidu will be contributing a version of its fault tolerable open source deep learning platform PaddlePaddle that integrates its Elastic Deep Learning feature for enhanced Kubernetes elastic scheduling. The Angel project will be contributed by Tencent.

"The Angel project, jointly developed by Tencent and Peking University, is a high-performance distributed machine-learning platform," Yongsheng Liu, general manager of Tencent's AI department, explained in a statement. "Tuned for big data/models, the Angel project is capable of supporting over a billion parameters."

Although AI has been a buzzword for over a decade, it's only been recently that it's been finding it's way into data center operations as well as commercial and consumer applications. In recent years the technology, especially machine learning, has shown a dramatic increase in use as the technology has matured, and today's users of commercial IT would be hard pressed to find applications that don't have some degree of machine learning built-in.

It's also a technology particularly suited for the open source development model when using permissive licensing, as companies can work together to build the basic AI or ML engines or frameworks and then add their own secret sauces in front of proprietary releases.

"Nokia considers AI as a game-changer for our customers," Henri Tervonen, CTO of Nokia's mobile networks, said in a statement. "We are already seeing breakthrough results from AI/ML in our networks and service offerings. Nokia is looking forward to supporting Acumos and influencing the platform’s evolving architecture."

Hide comments

Comments

  • Allowed HTML tags: <em> <strong> <blockquote> <br> <p>

Plain text

  • No HTML tags allowed.
  • Web page addresses and e-mail addresses turn into links automatically.
  • Lines and paragraphs break automatically.
Publish